Innovations in medical imaging:

Transcription

1 Innovations in medical imaging: research in academics and industry Norbert J. Pelc, Sc.D. Departments of Bioengineering and Radiology Stanford University Acknowledgements Brian Hargreaves, Stanford University Jiang Hsieh, GE Healthcare Kim Butts Pauly, Stanford University Robert Herfkens, Stanford University Research support: GE Healthcare Philips Healthcare Samsung Electronics Outline Impact of medical imaging (medical, economic) Progress through innovation in diagnostic imaging Vignettes during my time in industry Vignettes from academia Conclusions

2 Economic impact annual worldwide market 2014 CT $4B MRI PET & SPECT $5.5B $9.5B Ultrasound $6B TOTAL $27B Sources: BBC Research, MarketsandMarkets, Medical News Today, Transparency Market Research Economic impact annual worldwide market CT $4B ~0 MRI $5.5B ~0 PET & SPECT $9.5B ~0 Ultrasound $6B ~0 TOTAL $27B ~0 Sources: BBC Research, MarketsandMarkets, Medical News Today, Transparency Market Research Economic impact ~75M CT and ~35M MRI procedures/yr in USA Medical imaging is ~$100B per year (US) Lots of people employed Many patients are impacted Sources: Healthday, Organization for Economic Co-operation and Development

3 Medical impact Pnumoencephalography no longer used (A) AP and (B) lateral pneumoencephalography of patient with congenital toxoplasmosis with hyrocephalus. Medical impact modern imaging replaced most of exploratory surgery Medical impact patient with meningioma

4 Medical Impact How do innovations have impact? Publications and dissemination can alter research, and clinical use if changes don't require system mods or approvals Broad impact of technological changes requires commercialization Progress through innovation EMI Mark I 1973

5 Progress through innovation Radiology 124:81-86, Progress through innovation 1976 now Progress through innovation 1983 now RadioGraphics 4:5-25,

6 Progress through innovation Completely new product: commercialization is often by a new company Improvement to existing product: commercialization is usually through existing company Outline Impact of medical imaging Progress through innovation in diagnostic imaging Vignettes during my time in industry Vignettes from academia Conclusions Improved calibration methods new methods improved reliability and image quality and reduced manufacturing and service cost

7 Quarter-detector offset Hsieh, "Computed Tomography Principles, Design, Artifacts, and Recent Advances", "Bone-detail" reconstruction standard bone detail Hsieh, "Computed Tomography Principles, Design, Artifacts, and Recent Advances", "Bone-detail" reconstruction standard bone-detail Images courtesy of Jiang Hsieh

8 resp position resp position resp position resp position Observations Problems were important to the business There was an in-house expert and advocate (me) Relatively easy implementation Incorporated into the product quickly Innovations helped the customers and the company Respiration artifacts in MRI phase encoding repetition number 0.5 Conventional Acquisition Result => ghosting k y Respiratory Compensation phase encoding repetition number k y Result => reduced ghosting

9 Respiration artifacts in MRI Conventional Acquisition Respiratory Compensation Observations Feature was important to the business There was an in-house expert and advocate (me) Complicated implementation: hardware, operator interface, scan software, pulse sequences, reconstruction, operator interface Incorporated into the product Cardiac pulsation artifacts Zhu J, Gullapalli RP. "AAPM/RSNA Physics Tutorial for Residents: MR Artifacts, Safety, and Quality Control." RadioGraphics 2006; 26: Potential solution: phase encoding order based on cardiac cycle Relatively simple adaptation of resp comp feature already in the product Stanford clinicians liked it Not incorporated in the product

10 Observations Moderately easy implementation Not incorporated into the product Lack of perceived market impact? No in-house expert/advocate? Needed resources were required elsewhere? MRI near metal Four common metal artifacts are shown Signal Loss Failed Fat Suppression Distortion / Displacement Hyperintense signal pile-up Often occur together, from multiple mechanisms courtesy of Brian Hargeaves, PhD Frequency Bands Near Metal SEMAC Slice Encoding for Metal Artifact Correction cord impingement? Selective Slices Near Metal Combination of techniques - longer scan time but more robust in the presence of metal Developed at Stanford but GE was informed regularly SEMAC Excitation / Imaging 512x192, 3 NEX, 3mm T1W Spin Echo (5min) z FOV: 256x192, Encompass 60% ky, 4mm PD Off-Resonance SEMAC (11 min) Lu, et al. MRM 2009 courtesy of Brian Hargeaves, PhD Stanford University Radiology

11 MAVRIC Frequency Bands Near Metal MAVRIC Excitation / Imaging Multiacquisition Variable-resonance Image Combination Developed independently by a scientist at GE (Kevin Koch) z FOV: Encompass Object Different combination of techniques, not slice selective Koch, et al. MRM 2009 courtesy of Brian Hargeaves, PhD Stanford University Radiology MAVRIC-SL Kevin Koch and Brian Hargreaves knew and understood each other's techniques. Wide and open communications SEMAC and MAVRIC have different strengths and weaknesses. One is superior in some cases; the other is others A "hybrid of the two was created and implemented in GE product (MAVRIC-SL) Methods published and licensed. Similar techniques implemented by other vendors MAVRIC-SL Courtesy of Drs. Weitian Chen and Kevin Koch, GE courtesy of Brian Hargeaves, PhD Stanford University Radiology Observations Moderately complicated implementation Feature was important to customers There was an in-house expert and advocate Incorporated into the product Communication between the academic and industry groups, before and after commercialization decision, was critically important

12 Summary Medical imaging innovations have had a huge impact, clinically and economically. Medical physicists contributed greatly Impact of innovations needs dissemination. Commercialization makes innovations widely available to the medical community and public Summary Physicists within companies play a great role - innovators - experts and advocates for innovations internal innovations (including their own) external advances Physicists in the community and academia are closer to the clinician and clinical needs Dissemination decisions Commercialization decisions are complex - Assessment of commercial opportunity - Needed resources? Are they available? Within companies: internal matter. Physicists should be engaged.

13 Dissemination decisions Commercialization/dissemination of academic innovations can be by several routes: - Place it in the public domain - Patent and license. Expectations must be realistic. - Starting a new company may be appropriate for completely new product - Working with existing company may be more effective for evolutionary advances Academic - industry partnership "Evolutionary" advances accumulate to huge impact Collaboration and communication between academia and industry is essential New compliance policies and expectations of both sides can be a barrier Don't throw out the baby with the bathwater Thank You